dependency version updates (#324)

Author: themattinthehatt

docs/requirements.txt

@@ -7,13 +7,13 @@ sphinx-design
 aniposelib
 h5py
 hydra-core
-imgaug
+imaug
 kaleido
 kornia
 lightning
 matplotlib
-moviepy<2.0.0
-numpy<2.0.0
+moviepy>=2.0.0
+numpy>=2.0.0
 opencv-python
 pandas
 pillow

lightning_pose/utils/cropzoom.py

@@ -5,7 +5,7 @@
 import numpy as np
 import pandas as pd
 import tqdm
-from moviepy.editor import VideoFileClip
+from moviepy import VideoFileClip
 from omegaconf import DictConfig
 from PIL import Image
 from typeguard import typechecked
@@ -210,8 +210,7 @@ def crop_frame(get_frame, t):
 
         return cv2.resize(cropped_frame, (w, h))
 
-    # renamed image_transform in 2.0.0
-    cropped_clip = clip.fl(crop_frame, apply_to="mask")
+    cropped_clip = clip.transform(crop_frame, apply_to="mask")
 
     cropped_clip.write_videofile(str(output_file), codec="libx264")
 

lightning_pose/utils/predictions.py

@@ -15,7 +15,7 @@
 import numpy as np
 import pandas as pd
 import torch
-from moviepy.editor import VideoFileClip
+from moviepy import VideoFileClip
 from omegaconf import DictConfig, OmegaConf
 from torchtyping import TensorType
 from typeguard import typechecked
@@ -703,7 +703,7 @@ def create_labeled_video(
         upsample_factor = 1
 
     if upsample_factor > 1:
-        clip = clip.resize((upsample_factor * nx, upsample_factor * ny))
+        clip = clip.resized((upsample_factor * nx, upsample_factor * ny))
         nx, ny = clip.size
 
     print(f"Duration of video [s]: {np.round(dur, 2)}, recorded at {np.round(fps_og, 2)} fps!")
@@ -783,7 +783,7 @@ def add_marker_and_timestamps(get_frame, t):
         )
         return frame
 
-    clip_marked = clip.fl(add_marker_and_timestamps)
+    clip_marked = clip.transform(add_marker_and_timestamps)
     clip_marked.write_videofile(
         output_video_path, codec="libx264", fps=fps or fps_og or 20.0
     )

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "poetry.core.masonry.api"
 
 [project]
 name = "lightning-pose"
-version = "2.0.0"
+version = "2.0.1"
 description = "Semi-supervised pose estimation using pytorch lightning"
 license = "MIT"
 readme = "README.md"
@@ -34,13 +34,13 @@ dependencies = [
     "aniposelib",
     "h5py",
     "hydra-core",
-    "imgaug",
+    "imaug",
     "kaleido",
     "kornia",
     "lightning",
     "matplotlib",
-    "moviepy (<2.0.0)",
-    "numpy (<2.0.0)",
+    "moviepy (>=2.0.0)",
+    "numpy (>=2.0.0)",
     "opencv-python-headless",
     "pandas (>=2.0.0)",
     "pillow",

tests/api/test_model.py

@@ -103,7 +103,8 @@ def test_predict_on_video_file_singleview(tmp_path, request, toy_data_dir):
 
     # Test labeled_video generation.
     model.predict_on_video_file(
-        Path(toy_data_dir) / "videos" / "test_vid.mp4", generate_labeled_video=True
+        Path(toy_data_dir) / "videos" / "test_vid.mp4",
+        generate_labeled_video=True,
     )
     assert (model.labeled_videos_dir() / "test_vid_labeled.mp4").is_file()
 
@@ -113,7 +114,8 @@ def test_predict_on_video_file_method_multiview_model(tmp_path, request, toy_mda
 
     # Test prediction on a test video.
     model.predict_on_video_file(
-        Path(toy_mdata_dir) / "videos" / "test_vid_top.mp4", generate_labeled_video=True
+        Path(toy_mdata_dir) / "videos" / "test_vid_top.mp4",
+        generate_labeled_video=True,
     )
     assert (model.video_preds_dir() / "test_vid_top.csv").is_file()
     assert (model.video_preds_dir() / "test_vid_top_temporal_norm.csv").is_file()